This revised project will test whether dendritic cells (DC) modified with pox vectors encoding the tumor antigen carcinoembryonic antigen (CEA) will induce a clinically-effective frequency of tumor antigen-specific T cells in patients with colon cancer. Antigen-specific T cell activation and proliferation are essential for clinically effective immune responses. Promising methods for activating antigen-specific T cells include the use of antigen loaded DC. During the initial period of funding, we administered DC loaded with mRNA encoding CEA to patients with minimal residual disease following resection of hepatic metastases of colon cancer. Although the DC were well-tolerated, we observed little augmentation of the CEA-specific immune response and no improvement in recurrence free survival. This can be explained partly by inadequate antigen expression and presentation within the DC, by poor T cell activation due to low levels of costimulatory molecules such as CD80 on the immature DC, and by limited immunogenicity of the native tumor antigen. To overcome these problems, we have modified DC with recombinant, replication-defective pox vectors (including vaccinia (rV) and fowlpox (rF)) that express a TRiad of COstimulatory Molecules (CD80, CD54, and CD58, designated TRICOM) as well as a modified tumor antigen CEA(6D), with an amino acid substitution that creates a more potent epitope. In a phase I study of patients with CEA-expressing malignancies who received ex vivo generated DC modified with rF-CEA(6D)-TRICOM, we detected antigen-specific T cell responses in the range of 0.1-0.5% of the peripheral blood mononuclear cells (PBMC). Clinical benefit was associated with the highest levels of immune response. Recently, it has been observed that greater levels of immune response to pox vector-encoded antigens occur with heterologous prime-boost immunizations. Priming with vaccinia encoding CEA followed by boosts with fowlpox encoding CEA has been suggested to improve survival in patients with advanced cancer. Nonetheless, the induced CEA-specific immune responses were 0.01% or less of the PBMC. We hypothesize that prime-boost strategies with pox vector-modified DC will achieve greater clinical benefit by increasing the level of CEA-specific T cell activity. Therefore, we propose a phase II study to choose between two immunization strategies, DC infected with rV-CEA(6D)-TRICOM followed by DC infected with rF-CEA(6D)-TRICOM or rV-CEA(6D)-TRICOM followed by rF-CEA(6D)-TRICOM, in terms of which is associated with a better rate of disease-free survival at 2 years following hepatic metastasis resection and adjuvant chemotherapy. We will also measure the rate and magnitude of the CEA-specific immune response by ELISPOT. IL-2 increases the magnitude of the CEA specific T cell responses to pox-vector immunizations. Therefore, we will subsequently explore the role of IL-2 in augmenting CEA-specific immunity when given with the optimal immunization strategy.